In the drive for higher integration and operating speeds in LSI devices, the pattern rule is made drastically finer. The rapid advance toward finer pattern rules is grounded on the development of a projection lens with an increased NA, a resist material with improved performance, and exposure light of a shorter wavelength. To the demand for a resist material with a higher resolution and sensitivity, acid-catalyzed chemical amplification positive working resist materials are effective as disclosed in U.S. Pat. No. 4,491,628 and U.S. Pat. No. 5,310,619 (JP-B 2-27660 and JP-A 63-27829). They now become predominant resist materials especially adapted for deep UV lithography.
Also, the change-over from i-line (365 nm) to shorter wavelength KrF laser (248 nm) brought about a significant innovation. Resist materials adapted for KrF excimer lasers enjoyed early use on the 0.30 micron process, went through the 0.25 micron rule, and currently entered the mass production phase on the 0.18 micron rule. Engineers have started investigation on the 0.10 micron rule, with the trend toward a finer pattern rule being accelerated.
For ArF laser (193 nm), it is expected to enable miniaturization of the design rule to 0.09 μm or less. Since conventionally used novolac resins and polyvinylphenol resins have very strong absorption in proximity to 193 nm, they cannot be used as the base resin for resists. To ensure transparency and dry etching resistance, some engineers investigated acrylic and alicyclic (typically cycloolefin) resins as disclosed in JP-A 9-73173, JP-A 10-10739, JP-A 9-230595 and WO 97/33198.
With respect to F2 laser (157 nm) which is expected to enable further miniaturization to 0.07 μm or less, more difficulty arises in insuring transparency because it was found that acrylic resins which are used as the base resin for ArF are not transmissive to light at all and those cycloolefin resins having carbonyl bonds have strong absorption. It was also found that poly(vinyl phenol) which is used as the base resin for KrF has a window for absorption in proximity to 160 nm, so the transmittance is somewhat improved, but far below the practical level.
Since carbonyl groups and carbon-to-carbon double bonds have strong absorption in proximity to 157 nm as mentioned above, reducing the number of such units is considered to be one effective way for improving transmittance in proximity to 157 nm. It was recently found that introducing fluorine atoms into base polymers makes a great contribution to an improvement in transparency in the F2 laser region. It was reported, for example, in SPIE 2001, Proceedings 4345, pp. 273–284, “Polymer design for 157 nm chemically amplified resists” that in resist compositions comprising a copolymer of tert-butyl α-trifluoromethylacrylate with 5-(2-hydroxy-2,2-bistrifluoromethyl)ethyl-2-norbornene and a copolymer of tert-butyl α-trifluoromethylacrylate with 4-(2-hydroxy-2,2-bistrifluoromethyl)methylstyrene, the absorbance of the polymer at 157 nm is improved to about 3. However, this resin is still insufficient in transparency because it is believed that an absorbance of 2 or less is necessary to form a rectangular pattern at a film thickness of at least 2,000 Å through F2 exposure.
In this regard, a highly transparent resin having an absorbance of up to 1 is described in SPIE 2002, Proceedings 4690, pp. 76–83, “Synthesis of novel fluoropolymers for 157 nm photoresists by cyclo-polymerization.” This polymer has not only high transparency, but also good substrate adherence and developer affinity. The drawback is its critically low resistance to dry etching as compared with general use polymers for KrF and ArF laser exposure.